Sheer et al. normalized subjects' sleep patterns over a two-week period and then forced them to live for seven days on a 28-hour-per-day schedule, which rotated them from complete synchronization of their sleep-wake cycle with their biological circadian cycle through complete de-synchronization back to synchronization. Different biometric measurements were taken througout the day. Results demonstrated that metabolic cycles did not follow the sleep-wake patterns, as demonstrated by this graph of leptin, showing how its levels were synchronized, then desynchronized and synchronized again.
Results of other measurements such as epinephrine, norepinephrine, glucose, etc. demonstrated essentially stressful and adverse reactions to the shift work.
Although my intuition from having worked nights for a while is that this research is pretty much on the money, there were a few possible problems. For example, subjects seem to have been sequestered from sunlight (so that they wouldn't know where in a 24-hour cycle they were), and they were given food rather than being allowed to eat what they wanted. Were they given tasks to perform during their "days"? It doesn't say. Most problematically, they weren't allowed to exercise. From what I can tell, the exercising night shift workers are able to deal with the stress of working night much more easily than the non-exercising workers.
Of particular interest to me from the results was the fact that de-synchronized shifts created pre-diabetic states in 30% of the subjects, but not all of them. It would be interesting to follow night shift workers now and see if similar epidemiological results could be found by controlling for weight and eating habits.
Anyhow, I think this report justifies the $1.25 per hour shift differential that I received on nights.
- F. A. J. L. Scheer, M. F. Hilton, C. S. Mantzoros, S. A. Shea (2009). Adverse metabolic and cardiovascular consequences of circadian misalignment Proceedings of the National Academy of Sciences DOI: 10.1073/pnas.0808180106